Usman Salis
Artificial intelligence is transforming the global economy at breakneck speed, but there's a hidden bottleneck: computing power. Each new AI model generation needs drastically more processing capacity than the last, creating explosive demand for chips, servers, and cloud services. This imbalance between AI's rapid evolution and the infrastructure needed to support it is reshaping entire industries.
The Computing Power Crunch
A recent observation from Shay Boloor captured this perfectly: every AI upgrade demands exponentially more compute, which explains the soaring demand for companies like Nvidia (NVDA), AMD (AMD), Oracle (ORCL), NBIS, IREN, and CRWV. The "intelligence economy" is scaling faster than anyone anticipated, and infrastructure is struggling to keep pace.
Modern AI systems, especially large language models, require billions or even trillions of parameters to function. Training and running these models takes cutting-edge GPUs, massive data centers, and optimized cloud infrastructure. This has turned hardware manufacturers and infrastructure providers into the backbone of the AI revolution. Nvidia leads the pack with its H100 and upcoming B100 chips that power most AI training operations. AMD is pushing hard with its MI300 series to capture market share. Oracle has emerged as a major AI cloud partner, going head-to-head with Microsoft Azure and AWS. Meanwhile, players like NBIS, IREN, and CRWV are racing to build out infrastructure for compute-heavy workloads.
When Demand Outpaces Supply
The core issue is simple but significant: AI adoption is moving faster than the physical and digital infrastructure can support it. This creates serious supply chain bottlenecks for GPUs and semiconductors, drives up cloud service costs as demand overwhelms capacity, and opens doors for new players in AI hardware, networking, and energy-efficient computing. As each AI model grows more complex, the market for compute solutions expands right alongside it.
If current trends hold, the biggest winners will be companies sitting at the intersection of AI and compute infrastructure. Nvidia and AMD benefit from chip dominance. Oracle and other cloud providers gain from infrastructure growth. Emerging specialists like IREN and CRWV profit from data center scaling. For investors, the real question isn't whether demand will rise—it's which companies can scale quickly enough to capture the opportunity.
Usman Salis
